基于虚拟化技术的仿真系统容错优化方法

发布时间：2018-04-27 01:35

本文选题：仿真系统 + 容错　；参考：《计算机应用》2015年08期

【摘要】：节点崩溃或者仿真资源不足导致的分布式仿真系统故障,降低了仿真系统可靠性。为保证系统容错效果,降低容错开销,提出了一种基于虚拟化技术的仿真系统容错方法,按照系统故障发生的位置,对不同类型故障动态采用不同类型的容错策略。分析了检查点容错策略的优化方法,给出了最优设置间隔;结合虚拟化技术的优势,解决了副本容错策略的节点选择、副本数量以及位置分布问题;同时,引入基于虚拟机迁移的容错策略,并将其作为检查点容错策略和副本容错策略的补充,以降低容错开销。通过仿真实验数据对比,分析了动态容错策略与普通容错策略的性能,可知动态容错策略保证了系统容错性能,容错开销也保持在较低水平。
[Abstract]:The failure of distributed simulation system caused by node crash or insufficient simulation resources reduces the reliability of simulation system. In order to ensure the fault-tolerant effect of the system and reduce the fault-tolerant overhead, a fault-tolerant method for simulation system based on virtualization technology is proposed. According to the location of the system fault, different types of fault-tolerant strategies are used dynamically for different types of faults. The optimization method of fault-tolerant policy of checkpoint is analyzed, the optimal setting interval is given, and the problems of node selection, replica number and location distribution of replica fault-tolerant policy are solved by combining the advantages of virtualization. A fault-tolerant strategy based on virtual machine migration is introduced and used as a complement to checkpoint fault-tolerant strategy and replica fault-tolerant strategy in order to reduce the fault-tolerant overhead. The performance of dynamic fault-tolerant strategy and common fault-tolerant strategy is analyzed by comparing the simulation data. It is shown that the dynamic fault-tolerant strategy ensures the fault-tolerant performance of the system and the fault-tolerant overhead is kept at a low level.
【作者单位】：军械工程学院电子与光学工程系;
【基金】：装备预研基金资助项目(9140A04030213JB34001)
【分类号】：TP391.9;TP302.8

【参考文献】